This is a single-centre, open-label, parallel, pragmatic, superiority randomized clinical trial. Patients were recruited from the Hepatobiliary and Pancreatic (HBP) surgery unit in a referral tertiary Hospital (Donostia University Hospital). Patients were eligible if they were planned to undergo elective HBP surgery, irrespective to the surgical indication, and they agreed to participate voluntarily in the study and therefore signed the provided written informed consent. All patients accepted to comply with the study protocol postoperative follow-up in outpatients’ clinics. There were no restrictions regarding surgical procedures. Exclusion criteria included: urgent or emergent surgery, class IV wounds according to the CDC classification (i.e. dirty or infected surgical site)(15) and the presence of an active uncontrolled intraabdominal infectious foci prior to the intervention. Additional exclusion criteria were intraoperative uncontrolled faecal contamination of the surgical site and patients requiring reintervention due to major postoperative complications during follow-up, considering the potential confounding factor with the primary endpoint. Furthermore, all patients not complying with the specific study perioperative SSI prevention bundle or if they refused to participate, either verbal or in writing, were also excluded.
All eligible patients provided written informed consent before undergoing study related procedures. The study was approved by the local Ethics Committee for Clinical Research of Donostia University Hospital and was conducted in accordance with the Declaration of Helsinki (World Medical Association, 2013). This trial was designed and reported according to the CONSORT guidelines(16) and was prospectively registered at ISRCTN registry, with the number ISRCTN 37315612 on the 14/01/2020.
Patients were recruited by participating staff surgeons and enrolled in the study prior to elective surgery. Randomisation was performed before elective surgery, once the operation date was scheduled, in a 1:1 allocation ratio in order to balance treatment over the following factors: active smoking or clinically relevant alcohol abuse, American Society of Anaesthesiologists (ASA) physical status classification ≥ 3, Body Mass Index (BMI) ≥ 25 Kg/m2, diagnosis of diabetes mellitus or chronic obstructive pulmonary disease requiring medical treatment and preoperative placement of a biliary drainage (i.e. either internal biliary prostheses or external-internal biliary drainage). All these variables had been already proven to be independent risk factors for si-SSI elsewhere in the literature(12, 13). Randomisation was performed by an external statistician using an online randomisation software. Equal probability of assignment to each study intervention was secured by the method of minimisation with a random element. Staff surgeons were notified of the treatment allocation by internal notification system once the scheduled surgical procedure was conducted. The investigators, surgeons, patients, and statistician were unmasked to the group the patient was randomly assigned to.
All patients must comply with a specific study protocol including the aforementioned bundle of preventive perioperative interventions for SSI: preoperative antibiotic prophylaxis following WHO specific recommendations(5), surgical-site preparation using alcohol-based antiseptic solution (2% chlorhexidine gluconate), use of dual ring wound protectors (Alexis®; Applied Medical), maintenance of adequate perioperative oxygenation along with normovolemia and normothermia, application of protocols for intensive perioperative blood glucose control when indicated, periodic replacement of surgical gloves (every 2–4 hours) along with change of new and sterile both surgical instruments and surgical drapes before closure; together with the use of antimicrobial-coated suture for abdominal wall closure (i.e. Triclosan coated polydioxanone; PDS-Plus®, Ethicon surgery) among other interventions. These have been validated in the literature and are nowadays strongly endorsed by WHO and main surgical societies including the American College of Surgeons and the Spanish Surgery Association(5, 17, 18).
In the control group skin staples were used for dermal closure, with intervals of 10-15mm. On the other hand, in the experimental group subcuticular suture was used, performing a continuous intradermal suture with 4/0 long-term absorbable monofilament material. In both groups abdominal fascia closure was performed with a continuous uninterrupted long-term absorbable suture (i.e., polydioxanone or polyglyconate) using “small-bites” technique. Subdermal layer (subcutaneous tissue) was closed when a depth ≥ 3cm was found. For standardisation of all the procedures, both perioperative SSI prevention measures and incision closure procedure; staff surgeons were requested to learn the described interventions by means of specific lectures and reading a particular study protocol document provided by the investigators team. Furthermore, all participating surgeons were instructed to identify si-SSI according to CDC criteria(1, 19). All patients must comply with the same perioperative study protocol, irrespective to the allocated study group.
The primary endpoint was the incidence of si-SSI, defined according to CDC definitions for SSI(1, 19) and must be confirmed by a staff surgeon.
A specific subgroup analysis was planned for the primary endpoint evaluating outcomes in the following subgroups: major hepatic resection, pancreatic resection, bilioenteric reconstruction and open or laparoscopic approach. The study secondary endpoints were overall major postoperative morbidity in both groups, defined as Clavien-Dindo classification class ≥ IIIa(20). Also, additional wound complications (i.e. seroma or dehiscence), median hospital length of stay and need for re-hospitalisation owing to si-SSI were analysed. A descriptive analysis of microbiological cultures and treatments for si-SSI was included. Presence of additional independent risk factors for si-SSI in elective HBP surgery was also explored.
All patients were followed-up for the first 30 postoperative days, for both primary and secondary outcomes. During immediate postoperative period daily physical examination of the patient was performed by attending surgeons of the study team. After hospital discharge, all patients were followed-up in outpatients’ clinic for the second and fourth postoperative weeks to ensure post-discharge si-SSI detection. Complementary tests included aseptically obtained wound exudate cultures. All diagnosed si-SSI were confirmed by the reference HBP staff surgeons of the study team. The study protocol was published and is available online at https://www.isrctn.com
Assuming an incidence of si-SSI in elective HBP surgery of 11% in staples group and 3% in subcuticular suture group(13), a 5% patient loss and a significance level of 95%, we calculated that a minimum of 318 patients were required to achieve a power of 80%. The expected recruitment period for the total study population was 16 months. However, due to major impact of SARS-CoV-2 pandemic on elective surgery procedures delay, as occurred worldwide, this period was extended to 26 months.
Data was collected prospectively by the investigators, and all patients were followed-up as per protocol using individual electronic patient records. Quantitative data is presented as mean (SD) or median (IQR). Qualitative data is presented as absolute numbers and proportions. We used an intention-to-treat (ITT) analysis for all outcomes, and per-protocol (PP) analysis was performed for the primary endpoint. Differences among both groups were evaluated using parametric and non-parametric tests, as corresponded. We assessed the primary endpoint using χ2 test and a confidence interval of differences was calculated. We analyzed secondary endpoints qualitative variables using χ2 test or Fishers’ exact test. Equally, we analyzed secondary endpoints quantitative variables using Students’ T test for independent samples, or its’ non-parametric corresponding Mann-Whitney U test. We evaluated normal distribution of study variables among both groups using Saphiro-Wilk test. Univariate analysis was performed using a single variable Cox proportional hazard model. Any variable achieving p < 0.2 in the univariate analysis was included in a multivariate Cox proportional hazards model with a backward elimination. All p values are two-tailed, and a p < 0.05 was considered statistically significant. We performed all statistical analysis using STATA v16.1 (StataCorp LLC, Texas (USA)).
The present study founders had no role in study design, data collection, data analysis, data interpretation, or writing the report and neither had access to study raw data. The corresponding author had final responsibility for the decision to submit for publication. All authors approved the final version of the manuscript submitted for publication.